Trade and industrial structure with large firms and heterogeneity

We develop a model of trade and firm heterogeneity in an oligopolistic setting. This setting generates key differences in terms of modelling setup, modelling predictions and welfare implications with respect to the existing literature on trade and firm heterogeneity. In terms of modelling setup our approach allows us to explore interaction between potentially large heterogeneous firms, in contrast to recent trade literature with heterogeneity and atomistic firms. As a result variables like market price and total sales vary endogenously as different firms enter the market. We offer a solution for the integer problem inherent in small group models, based on stochastic dominance. The model generates testable predictions deviating from the benchmark firm heterogeneity model of Melitz (2003) in terms of the effect of trade liberalisation on markups, market shares, the market price. We also derive predictions on the effect of distance and market size on the probability of zero trade flows and export prices. Our model features the possibility that welfare declines as a result of trade liberalisation. The result in Brander and Krugman (1983), the benchmark model for trade under oligopoly, that welfare unambiguously rises with free entry and might decline without free entry due to increased cross-hauling is reversed. In a setting with heterogeneous instead of homogeneous firms, welfare might decline with free entry. A negative welfare effect without free entry can be ruled out if the firm size distribution is sufficiently dispersed.

► Firm heterogeneity and trade is studied in oligopoly with a discrete number of firms. ► A solution for the integer problem is provided based on stochastic dominance. ► Predictions are derived on the effects of trade on markups, market size and price. ► Predictions are generated on the probability of zero trade flows and export prices. ► Trade might reduce welfare with free entry, but not without free entry.